Abstract
<p>Path planning is crucial for efficient utilisation of autonomous underwater vehicles. The goal of the mission of an autonomous underwater vehicle determines suitable strategies for path planning. Blind search methods can be used for off-line path planning for unknown environments to locate phenomena of interest. Different blind search patterns have been implemented and evaluated in terms of their ability to reach the mission’s goal. A novel blind search pattern that is based on a truncated Lévy distribution is also proposed and compared with other search patterns as path-planning algorithms. The simulations show that Lévy search pattern can outperform other search patterns for small size phenomena. On the other hand, the proposed inverse-Lévy pattern can locate large size phenomena more than other search patterns. The simulations show that the probability of locating the most important phenomenon by a single autonomous underwater vehicle using blind search patterns is much smaller than of a swarm of autonomous underwater vehicles in similar conditions. However, Lévy and inverse-Lévy can be used for the worst-case scenario of no communication nor ability to use feedback information.</p>